Strokes are the most commonly acquired neurological disability in adults and stroke rates are higher in women than in men. Affected women are more likely than men to live longer with their stroke-related disabilities. Robot-assisted therapy represents the cutting edge of stroke rehabilitation and promises to improve the lives of veterans with disabilities due to stroke. Preliminary studies using these tools have demonstrated compelling evidence of their effectiveness and limitations. However, inconsistent carryovers of these gains to ADLs are seen. Therefore, it is still not clear what treatment strategies maximize functional outcomes on ADLs. Understanding how to optimize robot-assisted therapies is important for enhanced patient rehabilitation and improved functional outcomes. Imaging techniques such as functional Magnetic Resonance Imaging (fMRI) and Diffusion Tensor Imaging (DTI) can assist us in determining who may benefit from robot-assisted training and why training induced functional cortical changes may occur after robot training. We propose to conduct an interventional study plus control to assess the effectiveness of three weeks of robot-assisted practice of tasks with skilled functional tasks that resemble ADLs. We will assess the ability of such training to effect immediate gains and long-term functional improvements. Further, using fMRI and DTI, we plan to associate behavioral changes seen in motor impairment levels and functional task performance levels with white matter injuries and connectivity and changes in oxygen utilization in the motor cortex as well as other areas of the brain. We hypothesize the following: Robot-assisted practice of skilled reaching tasks improves ADL function after stroke. Our short-term aims are to 1) assess short-term functional gains practice of these skilled tasks; 2) assess long-term (six months) maintenance of these improvements; 3) quantify the central nervous system changes associated with robotic training, and 4) identify trends across high and low responders in terms of patterns of change in cortical activity and type of white matter connectivity. Our long-term aims are to 1) optimize robot-training strategies for practice of ADLs for intensity and duration to maximize cortical changes and 2) determine how lesion characteristics affect changes seen in function, white matter connectivity, and cortical activity.